Creep-Fatigue Flaw Growth Analysis to Support Elevated Temperature Flaw Size Acceptance Criteria

Abstract

TerraPower, LLC is the Independent Consultant hired by ASME ST-LLC to analyze a matrix of typical elevated temperature components using recognized creep-fatigue flaw growth analysis methods and data.

The scope of the work is to calculate the largest initial flaw size for each case that satisfies the specified transient operating conditions: temperature, pressure, time, and cycles.

The analysis methodologies utilized are:

• American Petroleum Institute (API) 579-1/ASME Fitness-for-Service (FFS)-1, Part 10, including Annex F material models and data

• Electricite de France (EDF) Recommended Procedure R5 V4/5.

• Electric Power Research Institute (EPRI) Boiler Life Evaluation and Simulation System (BLESS) code, including embedded material models and data.

The operating duration is 200,000 hours (22.8 years).

Stresses from applied internal pressure, welding residual, and thermal conditions are considered.

Four components are considered: superheater tube, reheater tube, superheater pipe, and reheater pipe.

Three materials are considered: Grade 22, Grade 91 and SS 304H.

The flaws configurations analyzed varied in orientation: circumferential and longitudinal, in location: inside surface, outside surface, mid-wall, and geometry: infinite length/full circumferential and semi-elliptical.

Stresses through the thickness and time are calculated for each configuration via transient (for the ramp up and down conditions) and steady-state (for the hold times) finite element analyses.

Software is developed for the three analysis methods and the analyses are carried out in an iterative fashion, starting with an initial postulated value for the maximum allowable flaw and refining it via the convergence method of bisection.

The authors implemented these methodologies in the manner a typical user would, given the typically available resources and mechanical properties.

All assumptions and implementation choices are documented in Section 5.

Different implementation choices would result in more or less significantly different results.

The analyses are verified with the extensive use of unit-testing and specific hand calculations.

The results are shown in tabulated format in Section 6.

Several conclusions are drawn in Section 9.

The general observation is that the three methodologies, as implemented by the authors, can give significantly different results.

While the three methodologies give very similar, identical at times, results for fatigue crack growth, the methods vary in the way creep crack growth is calculated.

Overall, the R5 V4/5 methodology results in the highest initial allowable flaw sizes.

The R5 V4/5 and API 579-1 / ASME FFS-1 mechanics of creep crack growth offer very similar equations.

The main reason for the difference in results between the two methodologies is that the R5 V4/5 methodology specifies that the stress intensity factor and reference stresses are calculated only using primary stresses, while the a fitting function for the total stress distribution through the thickness is used for API 579-1 / ASME FFS-1 .

As a consequence, the creep crack growth rates are lower and the initial allowable flaw size is higher.